Research ArticleGEOPHYSICS

Repeating earthquakes record fault weakening and healing in areas of megathrust postseismic slip

See allHide authors and affiliations

Science Advances  05 Aug 2020:
Vol. 6, no. 32, eaaz9317
DOI: 10.1126/sciadv.aaz9317
  • Fig. 1 Tectonic setting and instrumentation.

    Map view showing the spatial distribution of the Nicoya Seismic Cycle Observatory and Volcanological and Seismological Observatory of Costa Rica (OVSICORI-UNA) seismic stations used in this study (orange triangles). The color contours illustrate the coseismic slip area of the 5 September 2012, Mw 7.6, Nicoya Peninsula earthquake, slip increments of 0.5 m (47). The yellow regions up-dip and down-dip of the plate interface highlight the accumulated slow slip (in mm) between 2007 and 2017 (26). The black dashed line represents the East Pacific Rise (EPR)–Cocos Nazca Spreading (CNS) center boundary. Stars show the epicentral location of the RE clusters analyzed in this study color coded by cluster identity (id). Adjacent focal mechanisms show the faulting geometry of the target event in each cluster. Table S1 shows the corresponding hypocenter location and origin time for the five families. The focal mechanism shows the fault geometry and epicentral location of the 2012 Nicoya Peninsula earthquake (25). INDI, Punta Indio broadband seismic station.

  • Fig. 2 Temporal distribution of the repeating families analyzed in this study.

    (A to E) Magnitude distribution with respect to the time of occurrence of the 5 September 2012, Mw 7.6 Nicoya Peninsula earthquake (red dashed lines). Repeating earthquake clusters (RECs) for each family are formed by a target event (stars) and several smaller-magnitude events or EGFs (circles). (F) Spatial distribution of the REC with respect to the short-term afterslip, indicated by the white contour lines (27), and the long-term afterslip of the Nicoya Peninsula earthquake (28), indicated by the colored map. The amount of slip for both the short- and long-term afterslip is indicated in meters. Note the spatial decorrelation between the areas of maximum long-term afterslip and the location of the repeating families.

  • Fig. 3 Temporal variability of RE source.

    Earthquake source spectra and time function for three of the RE families: (A) C16, (B) C32, and (C) C60 obtained through the deconvolution procedure (fig. S2). Each panel shows the comparison between the megathrust-affected target event (blue) in 2012 and the healed target event (green) in 2015. The red and black dashed lines indicate the best-fit Boatwright model for each event, and the circles correspond to the corner frequency for each event. Note that for the three families, the healed (green) event shows more power at frequencies between 6 and 18 Hz, and its corner frequency is more than twice the corner frequency of the 2012 event. Avg., average.

  • Fig. 4 Stress drop variability as a function of time.

    (A) Stress drop variability as a function of time with respect to the occurrence of the 5 September 2012, Mw 7.6 Nicoya Peninsula earthquake (black dashed line). Circles represent the target event for each of the RECs within the five families, color-coded by family id (see Materials and Methods for calculation of error bars). The gray bars mark periods of slow slip events from 2008 to 2017 (26), and the orange and green regions indicate the duration of the short- and long-term Nicoya earthquake afterslip (27, 28). Note that no repeaters were identified after 2016. (B) Stress drop for each family as a function of recurrence time (tr) from previous event (circles) and time delay of first event after the Mw 7.6 Nicoya Peninsula earthquake (stars). Note that stress drop increases with log (tr), consistent with laboratory experiments (2, 19, 20).

  • Fig. 5 Space-time evolution of the RE fault patches before and after the 5 September 2012, Mw 7.6 Nicoya Peninsula earthquake.

    The figure shows the time-dependent behavior of frictional heterogeneities within the fault area of REs during the (A) late interseismic period (before the occurrence of the mainshock), (B) some days following the occurrence of the large event, and (C) ~1 year after the mainshock. The white regions of the fault patch represent areas with velocity-strengthening frictional properties, continuously creeping and elastically loading the edges of the unstable sliding asperities; the black areas with velocity-weakening (seismic) frictional properties. Following the occurrence of the mainshock in 2012, the slip area increases (from R1 to R2), but the effective area of contact, the black regions on the plate interface elastically coupled, reduces, decreasing the amount of slip in the subsequent rupture. As loading rate increases because of afterslip following the main event, asperities with conditionally stable sliding properties experience transient embrittlement, a transition from aseismic (light gray) to seismic (black) slip mode. After ~1 year following the Nicoya earthquake, afterslip amplitude reduces and some of the unstable areas (black) within the conditionally stable (dark gray) patches switch slip mode and behave aseismically again. Slip area reduces (from R2 to R3), but the effective area of contact increases.

Supplementary Materials

  • Supplementary Materials

    Repeating earthquakes record fault weakening and healing in areas of megathrust postseismic slip

    E. J. Chaves, S. Y. Schwartz, R. E. Abercrombie

    Download Supplement

    This PDF file includes:

    • Figs. S1 to S3
    • Table S1
    • Scaling between recurrence time and seismic moment
    • References

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article